 08 Nov, 2016 1 commit


Øyvind "pippin" Kolås authored
Add a property for GIMP to set when the used transformation matrix would make the size of the output explode. ref bug #316479

 30 Nov, 2013 1 commit


Daniel Sabo authored

 06 Jun, 2013 1 commit


Michael Muré authored
Also, renamme the parameter from filter to sampler

 03 Jun, 2013 1 commit


Michael Muré authored

 18 Jun, 2012 1 commit


Michael Muré authored

 01 Apr, 2012 1 commit


Øyvind "pippin" Kolås authored
It does perspective transforms as well now :)

 06 Apr, 2011 1 commit


Jon Nordby authored
This makes GeglMatrix3/2 binary compatible with GimpMatrix2/3 and introspectable by GObject introspection. Note: This is an API/ABI break. https://bugzilla.gnome.org/show_bug.cgi?id=645817

 05 Apr, 2010 1 commit


Øyvind "pippin" Kolås authored
It is better to always recompute, rather than rely on the unreliable state of a variable.

 14 Feb, 2010 1 commit


Michael Natterer authored
gegl_affine_get_bounding_box() called OpAffine::create_matrix() on a local variable instead of on affine>matrix, leaving subsequent users of affine>matrix with an identity matrix. Also, cleanup up the entite create_matrix() atuff:  add a wrapper function in affine.c which does not g_assert() if the function is not set.  remove typedef OpAffineCreateMatrixFunc because its signature was broken.  adapt chant.h accordingly.  adapt all subclasses to use the fixed signature. Plus some random code cleanups in affine.c

 20 Nov, 2009 1 commit


Øyvind "pippin" Kolås authored
Resurrected GEGL multi threaded locks with locking at the tile level, the performance gained from multi threading is small if detectable, through the added overhead and lock contention. There still are outstanding paralellisation issues. There is a per tile mutex, per buffer (in >tile_backend) mutex and a per node mutex. You need to pass enablemt to configure/autogen to enable the multi threading support.

 28 Nov, 2008 1 commit


Øyvind Kolås authored
* operations/affine/matrix.c: * operations/affine/matrix.h: moved these,.. * operations/affine/Makefile.am: .. * gegl/Makefile.am: .. here, and made it be GeglMatrix3 instead of just Matrix3. * gegl/geglmatrix.c: * gegl/geglmatrix.h: * gegl/geglplugin.h: Fixed up affine op to use the new code. * operations/affine/affine.c: (op_affine_init), (get_source_matrix), (get_bounding_box), (detect), (get_required_for_output), (get_invalidated_by_change), (affine_generic), (process): * operations/affine/affine.h: * operations/affine/chant.h: * operations/affine/reflect.c: (create_matrix): * operations/affine/rotate.c: (create_matrix): * operations/affine/scale.c: (create_matrix): * operations/affine/shear.c: (create_matrix): * operations/affine/transform.c: (create_matrix): * operations/affine/translate.c: (create_matrix): svn path=/trunk/; revision=2811

 14 Jun, 2008 1 commit


Øyvind Kolås authored
new framework. The patch isn't closed yet, but the code for nearest and linear is stable now and the rest of the code delta is quite large. The patch has been adapted to not introduce new API and the cubic and lanczos filters are disabled due to artifacts. * gegl/buffer/geglbufferaccess.c: * gegl/buffer/geglbuffertypes.h: * gegl/buffer/geglbuffer.c: * gegl/buffer/geglbuffer.h: * gegl/buffer/geglsamplercubic.c: * gegl/buffer/geglsamplerlanczos.c: * gegl/buffer/geglsamplerlinear.c: * gegl/buffer/geglsamplernearest.c: * gegl/buffer/geglsampler.c: * gegl/buffer/geglsampler.h: * operations/affine/affine.c: * operations/affine/affine.h: svn path=/trunk/; revision=2442

 03 Feb, 2008 1 commit


Øyvind Kolås authored
* gegl/geglplugin.h: modified to include the headers for gegloperation and it's subclesses (either by #include or directly through an ifdef, not sure which direction to take yet) * gegl/geglchant.h: modified to only include geglplugin.h * gegl/gegloldchant.h: .. here as well. * operations/affine/affine.h: * operations/color/remap.c: * operations/io/load.c: * operations/meta/layer.c: * operations/render/text.c: * tools/operation_reference.c: modified to not include the gegloperationfoo.h's directly, but rely on them being made available by geglplugin.h svn path=/trunk/; revision=1950

 03 Jun, 2007 1 commit


Øyvind Kolås authored
* configure.in: added gegl/operation/Makefile to the list of files to create. * gegl/Makefile.am: added subdir. * gegl/operation/Makefile.am: new makefile. * gegl/gegloperation*.[ch]: moved to a operation subdir. A lot of other files just had the path of some #includes modified. svn path=/trunk/; revision=1537

 11 Feb, 2007 1 commit


Øyvind Kolås authored
that is what they actually are. * operations/Makefile.am: * operations/transform/Makefile.am: * operations/transform/affine.c: * operations/transform/affine.h: * operations/transform/chant.h: * operations/transform/interpolatecubic.c: * operations/transform/interpolatecubic.h: * operations/transform/interpolatelanczos.c: * operations/transform/interpolatelanczos.h: * operations/transform/linear.c: * operations/transform/linear.h: * operations/transform/matrix.c: * operations/transform/matrix.h: * operations/transform/module.c: * operations/transform/module.h: * operations/transform/nearest.c: * operations/transform/nearest.h: * operations/transform/reflect.c: * operations/transform/rotate.c: * operations/transform/scale.c: * operations/transform/shear.c: * operations/transform/translate.c: svn path=/trunk/; revision=1322

 01 Oct, 2006 1 commit


Kevin Cozens authored
20061001 Kevin Cozens <kcozens@cvs.gnome.org> Added slightly modified versions of the lanczos and cubic interpolate affine transformations written by Geert Jordaens. * operations/transform/affine.c: Added hooks for interpolate ops. * operations/transform/affine.h: Added lanczos_width to _OpAffine. * operations/transform/interpolatecubic.c * operations/transform/interpolatecubic.h * operations/transform/interpolatelanczos.c * operations/transform/interpolatelanczos.h: Added * operations/transform/linear.c: Removed braces left in by mistake after removal of debug code.

 09 Aug, 2006 1 commit


Øyvind Kolås authored
